Abstract
We recently reported that the dimensions of face space are more well refined for young than older adult faces (Short & Mondloch, 2013). In the current study, we examined two alternative ways in which young and older adult faces might be represented in the context of a norm-based coding model. According to one model, adults possess a single age-generic norm that codes for both young and older faces, with face age represented as a dimension or set of dimensions within face space. Alternatively, there may be separable prototypes for young and older faces, with the representation of older faces being less well refined. In Experiment 1, 40 young adults participated in an opposing aftereffects experiment in which young and older faces were distorted in opposite directions (compressed versus expanded) during adaptation. Before and after adaptation, participants indicated which member of ±20% same-identity face pairs looked more normal; half of the pairs were older faces and half were young. Following adaptation, adults' normality preferences simultaneously shifted in opposite directions for the two face ages, p < .001, providing evidence for age-contingent opposing aftereffects. In Experiment 2, we sought to confirm these findings by examining the extent to which aftereffects transfer across face age categories. Pre- and post-adaptation trials were identical to those in Experiment 1; however, during adaptation, participants (n = 80) were adapted to either compressed or expanded faces from a single age category (young/old). Aftereffects, though significantly greater than chance for both face ages, were larger for the face age that matched adaptation than for the face age that did not, p < .01, indicating partial transfer of aftereffects across age categories. Collectively, these results suggest that adults process young and older adult faces with regard to separable prototypes with some shared coding dimensions.
Meeting abstract presented at VSS 2015